The shot gathers of the Finnish Reflection Experiment 2001–2003 are available as a set of SEG-Y files. Each SEG-Y contains typically a maximum of 1.1 GB of field recordings or a bit more than 50 shots. Typically, the file sizes range between 800–1000 MB. Where shot points have been skipped, the SEG-Y size is smaller.
These files are quality-controlled, meaning that noise tests, accidental recordings and duplicate traces have been removed and all relevant trace headers are set. The shot gathers have been prepared at the University of Helsinki, Institute of Seismology, during the year 2016.
The observer's notes and pole coordinates are also available as quality-controlled versions. Typical flaws in the original observer's notes include typing errors, wrong or impossible numeric values, and noise tests being erroneously flagged as recordings. A comprehensive erratum is available for all lines and can be downloaded under the OpenFIRE table view.
The samples are in IBM floats. The sampling interval is 2 milliseconds and the number of samples per trace is 15 000.
For the original, demultiplexed and correlated field recordings from Specgeofizika, see Raw data.
Finnish Reflection Experiment is a 2-D crooked-line seismic survey conducted using 4 to 5 Vibroseis sources per shot point. The pole interval of the survey is 50 meters. The nominal shot point interval is 100 meters, and each source and receiver group effectively corresponds to a single source or receiver placed in the midpoint of two adjacent poles.
The recording geometry is split-spread (asymmetric at the end of lines) with 362 active channels separated by a nominal group interval of 50 meters. This results in a nominal fold of ~90 for the entire survey.
Naming of files
The naming of files follows the general pattern
where the last two numbers separated by a dash specify a range of shot poles (source points). The shot gathers have been packaged so that around 50 shots should be included in each SEG-Y.
For example, the first 49 shots of FIRE 3-1 (shot points 1–94) are included in the file Fire_3_1_1-100.sgy, the next 48 (shot points 104–200) in the file Fire_3_1_101-200.sgy and so on.
Textual and binary header
The 3200-byte textual header follows a general format and is encoded in EBCDIC. Each row ends in an asterisk in column 80, and there are no linefeeds.
C1 FIRE Finnish Reflection Experiment 2001-2003 *
C2 Client : FIRE consortium: University of Helsinki, University of Oulu *
C3 and Geological Survey of Finland *
C4 Contractor: Spetsgeofyzika, Russia *
C5 LINE <NO> <LOCATION> ; acquisition: <DATE OF ACQUISITION> *
C6 Source: Vibroseis: 4-5 vibrators, nominal shot point interval 100 m *
C7 Vibrator group length: 50 m *
C8 Sweep length 30 sec, linear upsweep 12-80 Hz *
C9 Fully correlated trace length 30 sec (Total listening time 60 sec) *
C10 Vertical stacking: 8 sweeps / shot point, correlation before stack *
C11 Recording geometry: Split spread, asymmetric at the end of lines *
C12 362 active channels, group interval 50 m *
C13 Geophone groups: 12 geophones, group length 50 m *
C14 Number of samples: <SAMPLES> Sampling interval: <SAMPLERATE> *
C15 This reel contains the following field records (see obslogs): *
C16 <FLDR INTERVAL> *
C17 Number of traces in this reel: <NO TRACES> *
C18 Sourcepoint range: <EP INTERVAL> *
C19 CDP range: <CDP INTERVAL> *
C21 NOTE! trace header has the following non-standard fields: *
C22 bytes 171-172 (grnors): pole of present detector *
C23 bytes 173-174 (grnofr): pole of first detector in spread *
C24 bytes 175-176 (grnlof): pole of last detector in spread *
C26 <MISCELLANEOUS NOTES> *
C40 END TEXTUAL HEADER *
In the 400-byte binary header, the following fields are set:
- Bytes 13–14 (number of data traces per record), here 362 or 410
- Bytes 17–18 (sample interval in microseconds), here 2000 us
- Bytes 21–22 (number of samples per data trace), here 15 000
- Bytes 23–24 (number of samples per data trace for the original field recording), same as above
- Bytes 25–26 (data sample format code), here 1 (4-byte IBM floats)
- Bytes 27–28 (nominal CDP fold), here 90
- Bytes 29–30 (trace sorting code), here 1 (no sorting)
- Bytes 31–32 (vertical sum code), here 8
- Bytes 33–34 (lower sweep frequency), here 12 Hz
- Bytes 35–36 (upper sweep frequency), here 80 Hz
- Bytes 37–38 (sweep length), here 30 000 ms
- Bytes 39–40 (sweep type code), here 1 (linear)
- Bytes 49–50 (correlated data traces), here 2 (yes)
- Bytes 51–52 (binary gain recovered), here 1 (yes)
- Bytes 53–54 (amplitude recovery method), here 1 (none)
- Bytes 55–56 (measurement system), here 1 (meters)
In the trace headers, set are the following fields:
- Bytes 1–4 (trace sequence number), running numbering
- Bytes 5–8, as above
- Bytes 9–12 (original field record number)
- Bytes 13–16 (channel number)
- Bytes 17–20 (shot pole number)
- Bytes 21–24 (CMP number), the one closest to the source-to-receiver midpoint
- Bytes 25–28 (trace number within the CMP bin), for sorting
- Bytes 29–30 (trace identifier), here 1 (seismic data)
- Bytes 31–32 (number of vertically summed traces), here 8
- Bytes 33–34 (number of horizontally summed traces), here 1
- Bytes 35–36 (data use), here 1 (production)
- Bytes 37–40 (source-to-receiver offset in meters)
- Bytes 41–44 (receiver group elevation), in meters (Baltic/Kronstad system)
- Bytes 45–48 (surface elevation at source), in meters (Baltic/Kronstad system)
- Bytes 53–56 (datum elevation), 120 meters
- Bytes 57–60, as above
- Bytes 73–76 (source northing)
- Bytes 77–80 (source easting)
- Bytes 81–84 (receiver northing)
- Bytes 85–88 (receiver easting)
- Bytes 89–90 (coordinate units), here 1 (length units)
- Bytes 91–92 (weathering velocity), 600 m/s, for computation of field statics
- Bytes 93–94 (subweathering velocity), 5500 m/s, for computation of field statics
- Bytes 99–100 (source static correction), in milliseconds
- Bytes 101–102 (receiver static correction), in milliseconds
- Bytes 103–104 (total static applied), zero
- Bytes 115–116 (number of samples per trace), here 15 000
- Bytes 117–118 (sample interval in microseconds), here 30 000
- Bytes 119–120 (gain type), here 1 (fixed)
- Bytes 125–126 (correlated), here 2 (yes)
- Bytes 127–128 (sweep frequency at start), here 12
- Bytes 129–130 (sweep frequency at end), here 80
- Bytes 131–132 (sweep length in milliseconds), here 30 000
- Bytes 133–134 (sweep type), here 1 (linear)
- Bytes 157–158 (year), the four-digit year of recording
- Bytes 159–160 (day of year), Julian day
- Bytes 161–162 (hour of day), timezone unknown
- Bytes 163–164 (minute of hour)
- Bytes 165–166 (second of minute)
- Bytes 171–172 (pole of present detector)
- Bytes 173–174 (pole of first detector in spread)
- Bytes 175–176 (pole of last detector in spread)
The field record number refers to a row in observer’s logs.
Syntax of observer's notes
The observer's notes have a number of comment lines in the beginning of the files (all starting with the symbol #) and then line by line the recording geometry and time for each shot.
The observer's notes have 9 columns, from left to right FLDR (field record identifier, bytes 9–12 in trace headers), FIRST (pole of first geophone in spread), LAST (pole of last geophone in spread), SHOT (shot pole), D (day), M (month), H (hour), MIN (minute) and S (second).
# Observer's notes for FIRE 3-1
# The numbers in columns 2,3,4 refer to peg numbers.
# Format string: "%8d %7d %7d %7d %3d %3d %3d %3d %3d\n"
# Fire 3-1 starts here
# FLDR FIRST LAST SHOT D M H MIN S
2 1 362 1 8 5 12 41 5
3 1 362 3 8 5 12 51 13
4 1 362 5 8 5 13 1 7
5 1 362 7 8 5 13 10 1
Even the quality controlled versions of the data may contain errors. A comprehensive erratum is available for all lines and should be always consulted before using the data.
No attempt has been made to remove or mute noisy traces.